522 1,4-Dioxane Sensitivity

[jtingson]

HI everyone,

I am currently trying to set up method 522 for analysis of 1,4-dioxane in SIM mode. The main issue I am having is high background for ion 88, the QI for 1,4-dioxane. How can I decrease the baseline of the 88 ion to get better detection at lower concentrations of my analyte?

[70 eV]

QI is a pretty garbage abbreviation when quant and qual are both Q's. I've seen coelutions on 58 but 88 is usually just a falling baseline. What column are you running?

[jtingson]

My mistake, 88 is the quantitative ion.
The column used is a 30mx0.25mmx1.4um thickness RXI-624Sil MS

[70 eV]

Are you seeing the same baseline issue in your blanks? Have you tried re-shooting on scan and seeing if anything pops up as a TIC? We're running a CP Select 624 CB with the same dimensions down to a 0.05 ug/L reporting limit.

[jtingson]

Yes even when running blanks in SIM mode, the 88 baseline is still high.Full scan blank runs have also shown the same falling baseline for the 88 ion. We hypothesized that it could just be a result of DCM causing the high background but even when keeping the oven at 30 C for 5 minutes (where DCM elutes out fully in blank DCM run) the high background is still present.

[70 eV]

Not sure if you're going to be able to escape the falling baseline. Our integrator consistently takes an extra chunk of baseline on low points but since its baked into the calibration the quant values are fine. Tried setting it up to go valley to valley for awhile but it would cut the IS wonky on high concentration samples so we ditched it.

[jtingson]

Not sure if you're going to be able to escape the falling baseline. Our integrator consistently takes an extra chunk of baseline on low points but since its baked into the calibration the quant values are fine. Tried setting it up to go valley to valley for awhile but it would cut the IS wonky on high concentration samples so we ditched it.

What is the set up for you SIM? Dwell times, starts of segments, and gain factor?

[70 eV]

We run a half baked version of 522 since it never came under federal drinking water regulations. It's just 1,4-D and d-1,4-D as the IS. 4 ions total, single group, 100 ms each. SIM gain of 2, ~1450 volts and the multiplier is at least three years old. Tunes around 1250 and only cleaning it sees is the yearly PM. Runs ~800 samples a year so most of the time it's sittin around big chillin.

[jtingson]

This is currently what 2ppb standard looks like with an EM Voltage at 1525. We need to be able to detect 0.3 ug/L. What can I adjust to get this peak to be more prominent/ decrease the baseline noise.

https://www.dropbox.com/scl/fi/uxmvb8pv ... vq8rx&dl=0

[70 eV]

Our low point is 5 ppb, are you accounting for the 100x concentration factor that a sample would have? That would put you at 0.02 ppb.

[jtingson]

Appears using a 2mm liner seem to have fixed the signal issues being low instead of using 4mm liners. Signal looks to have improved significantly with this change.

[James_Ball]

Are you running splitless or a split mode injection?

Sometimes you can run split 1:5 and have almost the same sensitivity as splitless and lose a lot of the tailing of the solvent. Also if splitless, try shortening the splitless hold time until you see a decrease in the target peak, holding it too long will transfer more of the solvent to the column causing the tailing off that you see.

[bcd_GCLCMSMS]

Our low point is 5 ppb, are you accounting for the 100x concentration factor that a sample would have? That would put you at 0.02 ppb.

I believe the method (Option 2) is 100 mL sample, concentrated to 2 mL of eluent, so only a 50x concentration factor. Right?

We see a similar falling baseline.

Are your high baseline/backgrounds only observed in the extracted samples, or both the calibrators and extracted samples?

[jtingson]

QI is a pretty garbage abbreviation when quant and qual are both Q's. I've seen coelutions on 58 but 88 is usually just a falling baseline. What column are you running?

I have also seen coelutions of the 58 ion at low levels of 1,4-dioxane (<0.40ug/L) detection. This is something I only see in extracted samples (SIM mode) and causes quite a problem when it comes to identification of the analyte via EPA 522 method. What solutions can I take to separate the large 58 ion peak that coelutes with my analyte? (Currently using Rxi-624 Sil MS 30mx0.32mmx1.8um column) I have tried adjusting the temperature program as well but that has not yielded any promising results.

[bcd_GCLCMSMS]

We haven't had any luck resolving the M+ 58 peak either, although we can get down to 0.1 ug/L for 1,4-dioxane before the ion ratio integration falls outside of the method spec.

(Previously, this had been consistently <0.04 ug/L in our lab for the 100 mL extraction option before something in the supplies changed). All of our solvent blanks and unextracted blanks are still clean. A large bolus of the unknown contaminant comes off when first wetting the carbon cartridge, then it seems clean until the methanol is in the extraction and then more comes elutes off.